The regulation of body fat in mammals is a complex process involving the regulation of not only appetite but also energy expenditure. An important component of energy expenditure is non-shivering thermogenesis (NST). In rodents, the majority of NST appears to occur in brown adipose tissue (BAT) via the uncoupling protein (UCP) (Cannon & Nedergaard, 1985, Essays in Biochem. 20:110-165; Himms-Hagen J., 1989, Prog. Lipid Res. 28:67-115). UCP is a proton channel located exclusively in the inner mitochondrial membrane of adipocytes of the BAT (Nicholls & Locke, 1984, Physiol. Rev. 64:1-64). By allowing protons to equilibrate across the inner mitochondrial membrane, UCP uncouples oxidative phosphorylation from ATP production and thus converts stored energy into heat rather than work (Klingenberg M., 1990, Trends Biochem. Sci. 15:108-112; Klaus S. et al., 1991, Int. J. Biochem. 23:791-801). UCP-mediated uncoupling is not only capable of increasing body temperature in cold-acclimatized rodents and hibernating animals, but can also dissipate surplus caloric energy (Rothwell & Stock, 1986, In Brown Adipose Tissue. Trayhurn P., Nicholls D. G., Eds., London, Arnold, p. 269-298; Spiegelman & Flier, 1996, Cell 87:377-389; Hamann & Flier, 1996, Endocrinology 137:2129). A number of studies have now implicated UCP and brown adipose tissue as important regulators of body weight in rodents (Hamann & Flier, 1996, Endocrinology 137:2129; Lowell B. B. et al., 1993, Nature 366:740-742; Kopecky J. et al., 1995, J. Clin. Invest. 96:2914-2923; Cummings D. E. et al., 1996, Nature 382:622-626).
In humans, body weight homeostasis is poorly understood, but is also thought to involve regulated thermogenesis (Rothwell & Stock, 1981, Annu. Rev. Nutr. 1:235-56; Segal K. R. et al., 1992, J. Clin. Invest. 89:824-833; Jensen M. D. et al., 1995, Am. J. Physiol. 268:E433-438). However, the importance of the UCP in adult humans is questionable due to the low levels of BAT and consequently the low levels of UCP expression (Huttunen P. et al., 1981, Eur. J. Appl. Physiol. 46:339-345; Cunningham S. et al., 1985, Clin. Sci. 69:343-348; Schulz L., 1987, J. Am. Diet Assoc. 87:761-764; Santos G. C. et al., 1992, Arch. Pathol. Lab Med. 116:1152-1154).
In adult humans and other animals that do not contain large amounts of BAT, a large portion of NST and regulated thermogenesis is thought to be mediated by muscle and the white adipose tissue (Jensen M. D. et al., 1995, Am. J. Physiol. 268:E433-438; Davis T. R. A., 1963, Am. J. Physiol. 213:1423-1426; Astrup A. et al., 1989, Am. J. Physiol. 257:E340-E345, 1989; Simonsen L. et al., 1992, Am. J. Physiol. 263:E850-E855; Simonsen J. et al., 1993, Int. J. Obes. Relat. Metab. Disord. 17 (Suppl. 3):S47-51; Duchamp C. et al., 1993, Am. J. Physiol. 265:R1076-1083), however, the molecular mediators for regulated thermogenesis are currently unknown (Block B A., 1994, Annu. Rev. Physiol. 56:535-577).
Further, body weight disorders, including eating and other disorders affecting regulation of body fat, represent major health problems in all industrialized countries. Obesity, the most prevalent of eating disorders, for example, is the most important nutritional disorder in the western world, with estimates of its prevalence ranging from 30% to 50% within the middle-aged population. Other body weight disorders, such as anorexia nervosa and bulimia nervosa which together affect approximately 0.2% of the female population of the western world, also pose serious health threats. Further, such disorders as anorexia and cachexia (wasting) are also prominent features of other diseases such as cancer, cystic fibrosis, and AIDS.
Obesity, defined as an excess of body fat relative to lean body mass, also contributes to other diseases. For example, this disorder is responsible for increased incidences of diseases such as coronary artery disease, stroke, and diabetes. Obesity is not merely a behavioral problem, i.e., the result of voluntary hyperphagia. Rather, the differential body composition observed between obese and normal subjects results from differences in both metabolism and neurologic/metabolic interactions. These differences seem to be, to some extent, due to differences in gene expression, and/or level of gene products or activity. The nature, however, of the genetic factors which control body composition are unknown, and attempts to identify molecules involved in such control have generally been empiric and the parameters of body composition and/or substrate flux are monitored have not yet been identified (Friedman, J. M. et al., 1991, Mammalian Gene 1:130-144).
The epidemiology of obesity strongly shows that the disorder exhibits inherited characteristics, (Stunkard, 1990, N. Eng. J. Med. 322:1483). Moll et al., have reported that, in many populations, obesity seems to be controlled by a few genetic loci, (Moll et al. 1991, Am. J. Hum. Gen. 49:1243). In addition, human twin studies strongly suggest a substantial genetic basis in the control of body weight, with estimates of heritability of 80-90% (Simopoulos, A. P. & Childs B., eds., 1989, in "Genetic Variation and Nutrition in Obesity", World Review of Nutrition and Diabetes 63, S. Karger, Basel, Switzerland; Borjeson, M., 1976, Acta. Paediatr. Scand. 65:279-287).
Further, studies of non-obese persons who deliberately attempted to gain weight by systematically over-eating were found to be more resistant to such weight gain and able to maintain an elevated weight only by very high caloric intake. In contrast, spontaneously obese individuals are able to maintain their status with normal or only moderately elevated caloric intake.
In addition, it is a commonplace experience in animal husbandry that different strains of swine, cattle, etc., have different predispositions to obesity. Studies of the genetics of human obesity and of models of animal obesity demonstrate that obesity results from complex defective regulation of both food intake, food induced energy expenditure and of the balance between lipid and lean body anabolism.
There are a number of genetic diseases in man and other species which feature obesity among their more prominent symptoms, along with, frequently, dysmorphic features and mental retardation. Although no mammalian gene associated with an obesity syndrome has yet been characterized in molecular terms, a number of such diseases exist in humans. For example, Prader-Willi syndrome (PWS) affects approximately 1 in 20,000 live births, and involves poor neonatal muscle tone, facial and genital deformities, and generally obesity. The genetics of PWS are very complex, involving, for example, genetic imprinting, in which development of the disease seems to depend upon which parent contributes the abnormal PWS allele. In approximately half of all PWS patients, however, a deletion on the long arm of chromosome 11 is visible, making the imprinting aspect of the disease difficult to reconcile. Given the various symptoms generated, it seems likely that the PWS gene product may be required for normal brain function, and may, therefore, not be directly involved in adipose tissue metabolism.
In addition to PWS, many other pleiotropic syndromes which include obesity as a symptom have been characterized. These syndromes are more genetically straightforward, and appear to involve autosomal recessive alleles. The diseases, which include, among others, Ahlstroem, Carpenter, Bardet-Biedl, Cohen, and Morgagni-Stewart-Monel Syndromes.
Animals having mutations which lead to syndromes that include obesity symptoms have also been identified. Attempts have been made to utilize such animals as models for the study of obesity. The best studied animal models for genetic obesity are mice which contain the autosomal recessive mutations ob/ob (obese) and db/db (diabetes). These mutations are on chromosomes 6 and 4, respectively, but lead to clinically similar pictures of obesity, evident starting at about 1 month of age, which include hyperphagia, severe abnormalities in glucose and insulin metabolism, very poor thermo-regulation and non-shivering thermogenesis, and extreme torpor and underdevelopment of the lean body mass. Restriction of the diet of these animals to restore a more normal body fat mass to lean body mass ration is fatal and does not result in a normal habitus.
Although the phenotypes of db/db and ob/ob mice are similar, the lesions are distinguishable by means of parabiosis. The feeding of normal mice and, putatively, all mammals, is regulated by satiety factors. The ob/ob mice are apparently unable to express the satiety factor, while the db/db mouse is unresponsive to it.
In addition to ob and db, several other single gene mutations resulting in obesity in mice have been identified. These include the yellow mutation at the agouti locus, which causes a pleiotropic syndrome which causes moderate adult onset obesity, a yellow coat color, and a high incidence of tumor formation (Herberg, L. and Coleman, D. L., 1977, Metabolism 26:59), and an abnormal anatomic distribution of body fat (Coleman, D. L., 1978, Diabetologia 14:141-148). Additionally, mutations at the fat and tubby loci cause moderately severe, maturity-onset obesity with somewhat milder abnormalities in glucose homeostasis than are observed in ob and db mice (Coleman, D. L., and Eicher, E. M., 1990, J. Heredity 81:424-427). Further, autosomal dominant mutations at the adipose locus of chromosome 7, have been shown to cause obesity.
Other animal models include fa/fa (fatty) rats, which bear many similarities to the ob/ob and db/db mice, discussed above. One difference is that, while fa/fa rats are very sensitive to cold, their capacity for non-shivering thermogenesis is normal. Torpor seems to play a larger part in the maintenance of obesity in fa/fa rats than in the mice mutants. In addition, inbred mouse strains such as NZO mice and Japanese KK mice are moderately obese. Certain hybrid mice, such as the Wellesley mouse, become spontaneously fat. Further, several desert rodents, such as the spiny mouse, do not become obese in their natural habitats, but do become so when fed on standard laboratory feed.
Animals which have been used as models for obesity have also been developed via physical or pharmacological methods. For example, bilateral lesions in the ventromedial hypothalamus (VMH) and ventrolateral hypothalamus (VLH) in the rat are associated, respectively, with hyperphagia and gross obesity and with aphagia and cachexia. Further, it has been demonstrated that feeding monosodium-glutamate (MSG) to new born mice also results in an obesity syndrome.
Attempts have been made to utilize such animal models in the study molecular causes of obesity. For example, adipsin, a murine serine protease with activity closely similar to human complement factor D, produced by adipocytes, has been found to be suppressed in ob/ob, db/db and MSG-induced obesity (Flier, 1987, Science 237:405). The suppression of adipsin precedes the onset of obesity in each model (Lowell, 1990, Endocrinology 126:1514). Further studies have mapped the locus of the defect in these models to activity of the adipsin promoter (Platt, 1989, Proc. Natl. Acad. Sci. USA 86:7490). Further, alterations have been found in the expression of neuro-transmitter peptides in the hypothalamus of the ob/ob mouse (Wilding, 1993, Endocrinology 132:1939), of glucose transporter proteins in islet .beta.-cells (Ohneda, 1993, Diabetes 42:1065) and of the levels of G-proteins (McFarlane-Anderson, 1992, Biochem. J. 282:15).
To date, no gene, in humans, has been found which is causative in the processes leading to obesity. Likewise, to date, no molecular mediator of regulated thermogenesis in humans has been identified. Given the importance of understanding body weight homeostasis and, further, given the severity and prevalence of disorders, including obesity, which affect body weight and body composition, there exists a great need for the systematic identification of genes involved in these processes and disorders.